Departamento de Produccio n Agraria, Universidad Pu blica de Navarra, Pamplona, Spain; 2 ECOSUR, AP 36, Tapachula 30700, Chiapas, Mexico;

Transcription

1 Biocontrol Science and Technology (2001) 11, 649± 662 Consequences of Interspeci c Competition on the Virulence and Genetic Composition of a Nucleopolyhedrovirus in Spodoptera frugiperda Larvae Parasitized by Chelonus insularis A. ESCRIBANO, 1 T. WILLIAMS, 2 D. GOULSON, 3 R. D. CAVE, 4 J. W. CHAPMAN 3 and P. CABALLERO 1 1 Departamento de Produccio n Agraria, Universidad Pu blica de Navarra, Pamplona, Spain; 2 ECOSUR, AP 36, Tapachula 30700, Chiapas, Mexico; 3 Biological Sciences, University of Southampton, Southampton, SO16 7PX, UK; 4 Escuela Agrõ  cola Panamericana, AP 93, El Zamorano, Honduras (Received for publication 14 August 2000; revised manuscript accepted 21 February 2001) Nucleopolyhedroviruse s (Baculoviridae) are virulent insect pathogens that generally show a high degree of host speciwcity and have recognized potential as biologica l insecticides. Whenever viruses are applied for pest control, a proportion of the infected insects will also be parasitized by hymenopteran or dipteran parasitoids and interspeciwc competition for host resources will occur; the severity of such competition is likely to be modulated to a large degree by the virulence of each type of parasite. We examined the impact of parasitism by the solitary egglarval endoparasitoi d Chelonus insularis (Hymenoptera: Braconidae) on the speed of kill of nucleopolyhedrovirus-infected Spodoptera frugiperda (Lepidoptera: Noctuidae) larvae and the pattern of host growth and virus production in infected and/or parasitized hosts. We also examined the evect of parasitism on the virulence, infectivity and genetic composition of serially passaged virus. Both parasitism and viral infection resulted in a marked reduction in host growth. When third instar larvae were dually parasitized and virus-infected, the growth rate was even more severely avected compared to parasitized larvae. There was a signiwcant increase in virus production in larvae infected at later instars. InterspeciWc competition resulted in a substantial decrease in pathogen production in parasitized larvae infected at the fourth instar, but not in parasitized larvae infected at earlier instars. The serial passage experiment resulted in the appearance of four distinct genetic isolates of the virus detected by restriction endonuclease analysis. Of the three isolates that appeared in nonparasitize d larvae, two showed increased virulence, expressed by mean time to death, and for one of these the infectivity, expressed as LC 50, was reduced. One isolate that appeared in parasitized larvae (isolate D) had increased virulence and infectivity. Southern blot analysis indicated that virus isolate D Correspondence to: Dr. P. Caballero. Tel: ; Fax: ; pcm92@ unavarra.es. Present address for J.W. Chapman: Entomology and Nematology Department, IACR Rothamsted, Harpenden, Herts AL5 2JQ, UK. ISSN (print)/issn (online)/01/ DOI: / Taylor & Francis Ltd

2 650 A. ESCRIBANO ET AL. was most likely generated by point mutation of a restriction site or by alterations such as duplications, deletions or by recombination of two or more genotypic variants present in the wild-type nucleopolyhedroviru s isolate. Our study provides clear evidence of interspeciwc competition within the host, since, depending on the timing of inoculation, adverse evects were observed upon both the parasitoid and the virus. Keywords: Spodoptera frugiperda, Chelonus insularis, SfMNPV, infectivity, lethal time, genetic isolates, interspeciwc competition INTRODUCTION Insect baculoviruse s (Baculoviridae) are highly virulent pathogens that generally show a high degree of host speci city (Miller & Lu, 1997; Blissard et al., 2000). As such, they have attracted considerable attention as biological insecticides in programmes of integrated pest control (Moscardi, 1999). Due to their narrow host range, baculoviruse s are generally believed to be compatible with the action of other insect predators and parasitoids present in the crop habitat (Vasconcelos et al., 1996; Stark et al., 1999). Baculovirus isolates are normally a mixture of strains that diver in their virulence (Maruniak et al., 1999; MunÄ oz et al., 1999). In addition, virulence can diver according to the geographical origin of the virus (Shapiro et al., 1991; Escribano et al., 1999), or origin, species and growth stage of the host (Boots & Begon, 1995; Garcõ Â a-maruniak et al., 1996; Milks, 1997). Virulence can also be manipulate d arti cially by selection (Pavan & Ribeiro, 1989; Kolodny-Hirsc h & Van Beek, 1997) or by recombinant DNA technology employed to enhance the speed of kill of the virus (Bonning & Hammock, 1996). Although baculoviruse s do not infect insect natural enemies, an important aspect in the evaluatio n of baculoviru s bioinsecticides is the possible impact on predators and parasitoids through competitive or indirect evects (Brooks, 1993). In situations where a host is simultaneously infected by virus and parasitized by an insect parasitoid, possible interactions include the death of the parasitoid due to virus-induce d host mortality (Eller et al., 1988), or due to toxic factors produced by the virus in infected hosts (Hotchkin & Kaya, 1983). Virus production can also be impaired due to competition for host resources from the developing parasitoid (Hochberg, 1991a). Consequently, the degree of competition that occurs between baculoviruse s and parasitoids in dually infected and parasitized hosts is likely to be modulated to a large degree by the virulence of each type of parasite. Similarly, the pattern of pest mortality can be avected by changes in the susceptibility of parasitized hosts to viral infection (Beegle & Oatman, 1974; Santiago-Alvare z & Caballero, 1990). There are relatively few studies of the impact of virus infection on insect parasitoids (Brooks, 1993), and none, as far as we are aware, that examine the impact of the parasitoid on the genetic composition and phenotypic traits of viral progeny from dually infected hosts. The solitary egg-larval endoparasitoid Chelonus insularis (Cresson) (Hymenoptera: Braconidae) is one of the most abundan t parasitoids of the fall armyworm, Spodoptera frugiperda (Lepidoptera: Noctuidae) in the southern USA, Mexico and Central America (Ashley, 1986; Wheeler et al., 1989). The nucleopolyhedroviru s (NPV, Baculoviridae) of S. frugiperda has recognized potential as a biological insecticide (Moscardi, 1999) and is currently being developed for use by Mesoamerican maize growers (Williams et al., 1999). Whenever viruses are applied as bioinsecticides, a proportion of the infected hosts will also be parasitized by hymenopteran or dipteran parasitoids (Martõ Â nez et al., 2000). The impact of parasitism on viral insecticidal characteristics will be particularly relevant when virus production involve s the application of virus to eld population s of host larvae, followed by harvesting of virus-infected larvae and the subsequent recycling of this inoculum for pest control throughou t the growing season, as occurs in the control of Anticarsia gemmatalis, the major pest of soya in Brazil (Moscardi, 1989; Moscardi et al., 1997). In the present study, we examine the impact of parasitism on the speed of kill of virus-

3 NUCLEOPOLYHEDROVIRUS± PARASITOID COMPETITION 651 infected armyworm larvae and the pattern of host growth and virus production in infected and/or parasitized hosts. We also examine the evect of parasitism on the genetic composition and virulence of serially passaged virus in an attempt to elucidate the impact of parasitism on the virus infectivity and the virus virulence which are two key characteristics that avect the insecticidal potential of the virus. MATERIALS AND METHODS Biological Material Larvae of S. frugiperda were obtained from a culture established in 1997 at the Universidad Pu blica de Navarra, Spain, maintained on semisynthetic diet in a growth chamber at ë C, 85% relative humidity (RH), with a 16 h day length. The braconid egg-larval parasitoid Chelonus insularis originated from a culture started with eld-collected parasitized S. frugiperda larvae at El Zamorano, Honduras. C. insularis lay their eggs into eggs of S. frugiperda, but larval parasitoid development is completed much later when the parasitized host larvae begin metamorphosis in the fth instar while nonparasitized larvae do so in the sixth instar. Parasitized larvae included in the experiments detailed below were obtained by placing 400 recently laid eggs of S. frugiperda in a plastic Petri dish (115 mm diameter, 45 mm high), during 24 h, together with four C. insularis females that had been observed to mate 48 h previously. All experiments involved a multiple nucleopolyhedroviru s (MNPV) originally isolated from S. frugiperda larvae in Nicaragua and recently characterized (Escribano et al., 1999). Viral occlusion bodies (OBs) for use in experiments were extracted from dead diseased larvae by trituration in water and puri ed by various steps of diverential centrifugation as described elsewhere (Caballero et al., 1992). The concentration of the resulting puri ed suspension was determined using a bacterial counting chamber under phase-contrast microscopy. Mortality, Larval Growth and Virus Production Newly moulted third and fourth instar larvae, nonparasitized or parasitized by C. insularis, were allowed to drink from an aqueous suspension containing and OBs ml - 1 of virus, respectively, in a solution of 10% sucrose and 0.01% Fluorella blue dye. These concentrations represented the LC 9 0 values for third and fourth instar larvae, respectively, (Escribano et al., 1999). Control larvae, nonparasitized or parasitized by C. insularis, were treated with a solution of sterile water and food colouring. Thirty larvae that had completely ingested the solution within 10 min (as indicated by the blue coloration of the gut visible through the larval epidermis) were individuall y transferred into 25 ml cups containing diet and maintained at ë C, 85% RH, 16 h day length. Daily readings were taken on larval weight (6 0.1 mg) and mortality due to virus infection or parasitoid emergence. A further 20 second, third, four and fth instar larvae parasitized at the egg stage and nonparasitized larvae, were infected with their corresponding LC 9 0 (Escribano et al., 1999), and reared as above. Larvae that died of virus infection were individuall y transferred to a microfuge tube, and homogenized in 300 l l of distilled water. Viral OBs were puri ed according to the method of MunÄ oz et al. (1997) and yields of OBs per infected larva were determined by counting triplicate samples of diluted virus suspension under phase-contrast microscopy. In all cases, only intact insect cadavers were used to obtain virus yield data. The presence of a developing parasitoid in each S. frugiperda larva used in the parasitism treatments was con rmed by dissection under a binocular microscope. Serial Passage of Virus in Nonparasitize d and Parasitized Larvae Groups of 30 newly moulted third instar nonparasitized or parasitized larvae were infected as above with an LC 9 0 concentration of virus. Larvae were maintained on diet until they

4 652 A. ESCRIBANO ET AL. died or pupated. Virus-killed larvae from each treatment were collected and frozen. Viral OBs were recovered from each pool of larvae and puri ed as indicated above. A portion of the puri ed virus was used to infect other batches of nonparasitized and parasitized larvae, for the following passage. The remaining virus was used for viral DNA puri cation and bioassay analysis (see below). This method of in vivo passaging was repeated using groups of 30 third instar larvae for a total of 10 passages. The experiment was performed three times. DNA Analysis of Virus in Serial Passage Experiment Virions were released from viral OBs by incubation with 30 mm-na 2 CO 3, 50 mm-nacl, 30 mm-edta [ph 10.5] and puri ed by centrifugation in continuous sucrose gradient (Caballero et al., 1992). Puri ed virions were incubated with proteinase K (200 l g ml - 1 ) at 45± 50ë C for 2.5 h and then with 1% sodium dodecyl sulfate for an additional 30 min. After phenol-chlorofor m extraction, the aqueous suspension containing the DNA was dialyzed against three changes of 10 mm-te buver at 4ë C for 48 h. DNA was incubated with PstI or BglII (Amersham) at 37ë C for 4 h as described in the supplier s instructions. Reactions were stopped by the addition of loading buver (0.25% bromophenol blue, 40% sucrose). Electrophoresis was performed using 0.8% agarose gel in TAE buver (40 mm-tris-acetate, 1 mm-edta [ph 8.0]) containing 0.25 l g ml - 1 of ethidium bromide and photographed on a UV transilluminator. Southern Blot Hybridization For Southern blot analysis, the DNA from nonparasitized S. frugiperda larvae, parasitized S. frugiperda larvae, parasitized S. frugiperda larvae from which the parasitoid larvae had been removed, and adult female C. insularis was isolated using the procedure described by Heckel et al. (1995). DNA samples were digested with BglII, separated by electrophoresis and the gel was subjected to Southern blotting onto a Hybond-N + (Amersham) membrane overnight. A submolar DNA fragment that appeared in one DNA restriction pro le in virus recovered from larvae parasitized by C. insularis was used as a probe. This submolar fragment 3 was separated in low melting point agarose before being labelled with [a 2 2 P]dCTP by random priming (Sambrook et al., 1989). Labelled probe was allowed to hybridize at 42ë C in 5 3 SSC (1 3 SSC is 0.18 M NaCl, 10 mm-nah 2 PO 4, and 1 mm-edta, ph 7.7), 50% formamide, 20 mm-napo 4 (ph 7.0), 1 3 Denhardt s solution, and 0.1 mg of herring-sperm DNA. After hybridization, the blots were washed twice, 15 min each, at 42ë C in 2 3 SSC containing 0.1% SDS and exposed to X-ray lm (Biomax, Kodax, Rochester, N.Y.) for 3 to 48 h at 2 80ë C with an intensifying screen. Biological Activity of Virus in Serial Passage Experiment The biological activity of the virus isolates produced in the diverent passages was determined in terms of infectivity and virulence. The infectivity, de ned as the ability of a microorganism to produce infection, was expressed by means of LC 5 0, and the virulence, de ned as the disease producing power of a microorganism, was expressed by the mean time to death. The median lethal concentrations (LC 5 0 ) and the mean times to death of the original wild type isolate and the virus samples collected after 1, 3, 6 and 10 passages were determined by droplet bioassay (Hughes & Wood, 1981). First instar S. frugiperda larvae were selected at the moment they began to moult as determined by head capsule slippage. Selected larvae were starved for 8 h at 25ë C and were then overed an aqueous suspension containing virus at concentrations of , , , , or OBs ml - 1. This range of concentrations was found to kill between 5 and 95% of the test larvae in bioassays previously performed with the original isolate. Larvae that ingested the solution within 10 min were transferred to individua l cells of a 25-compartment plastic dish with diet and maintained at ë C. Larval mortality was recorded every 12 h until larvae had either died or pupated. Bioassays with 25 larvae per virus concentration plus 25 larvae as a control were performed three times.

5 NUCLEOPOLYHEDROVIRUS± PARASITOID COMPETITION 653 Data Analysis Virus yield and time-mortality data were analyzed for each host instar at time of virus infection and larval condition (nonparasitized or parasitized by C. insularis) using analysis of varianc e (ANOVA) models in the generalized linear modelling package of the program SPSS (v. 7.5). The data were obtained as one piece of information for each larva and, therefore, each data point was independent of the others. Model-checking showed that these data satis ed the assumptions of normality required by ANOVA. DiVerences between individua l treatments were assessed using independent contrasts. To perform these contrasts, data in diverent treatments were grouped together, and the old treatment variable was replaced in the ANOVA models by the new grouping. If this replacement caused a nonsigni cant change in the variation explained by the model, there was no signi cant diverence between the grouped treatments. Concentration-mortalit y regressions for the diverent virus inocula bioassayed were calculated with the probit option (Finney, 1971) of the computer program POLO-PC (Le Ora Software, 1987). RESULTS Mortality, Larval Growth and Virus Production Virus mortality for both nonparasitized and parasitized S. frugiperda larvae was in all cases greater than 95%. The mean time to death of virus-infected parasitized S. frugiperda larvae (98.4 h 3.33) did not diver signi cantly from that observed in virus-infected nonparasitize d larvae (100.8 h 4.61) (F 1, ; NS) of the third instar. In contrast, mortality due to parasitism in parasitized but not infected larvae occurred at h (6 5.8 h) after the start of the experiment in larvae parasitized by C. insularis (all gures are means 95% CL). Parasitism by C. insularis resulted in diminished host growth compared to the nonparasitized S. frugiperda larvae. Virus infection also severely reduced host growth from 48 h or 72 h onwards compared to mock-infected control larvae. Weight gain was most severely compromised in larvae that were dually virus infected and parasitized by C. insularis (Figure 1). There was a highly signi cant increase in virus production with larval instar (F 2, ; P < 0.001); the evect of parasitism was also highly signi cant (F 1, ; P< 0.001) and there was a signi cant interaction between these factors (F 2, ; P< 0.001). In nonparasitized larvae there was a signi cant positive relationship between virus yield and the stage of the larva at infection (Table 1). In parasitized larvae, the number of viral OBs produced in infected larvae was signi cantly reduced compared to nonparasitized larvae only in the fourth instar, not in the preceding instars. Since parasitoid emergence occurred prior to the fth instar, data on virus yield in parasitized fth instar larvae could not be obtained. DNA Analysis of Virus in Serial Passage Experiment In addition to the wild-type (WT) restriction endonuclease pro le described previously (Escribano et al., 1999), DNA extracted from viral OBs after passages 1, 3, 6 and 10 from virus-kille d nonparasitize d and parasitized larvae showed four isolates when subjected to restriction endonucleas e analysis using BglII or PstI (Figure 2). These isolates, hereafter referred to as A, B, C and D were characterized by the presence of a particular submolar fragment diverent for each isolate. Isolates A, B, and C, showed a submolar fragment at 6.1 kb, 8.1 kb, and 6.5 kb, respectively, when treated with PstI but were indistinguishabl e from the WT when treated with BglII. Isolate pro le D was characterized by a submolar fragment at 14.5 kb following treatment with BglII, but did not diver from the WT restriction endonucleas e pro le when treated with PstI. Pro le isolates A, B, and C were only detected in virus produced in nonparasitized S. frugiperda larvae while pro le isolate D was only detected in virus recovered from parasitized larvae (Table 2). Pro le isolate A was detected in experiments I and III at

6 654 A. ESCRIBANO ET AL. Mock Parasited(P) Infected(I) P+I a a Body weight (mg) a a b b b c a a c b b a b b c Hours postinfection Body weight (mg) b a a a b b b b a a c bc b c c b c Hours postinfection FIGURE 1. Body weight of Spodoptera frugiperda third (a) and fourth (b) instars infected with nucleopolyhedrovirus and/or parasitized by Chelonus insularis. Data were obtained as one piece of the information for each larva and timepoint and, therefore, each data point was independent of the others. Columns represent means 6 SE. passage 6 and 10, respectively, but in experiment I isolate A was replaced by isolate B at passage 10. Pro le C appeared in experiment II at passage 6 and remained unchanged until passage 10. Pro le D was detected in experiment II with larvae parasitized by C. insularis at passage 3. After appearing, this isolate remained unchanged in subsequent passages suggesting that it represents some consistent advantag e for the virus in parasitized hosts.

7 NUCLEOPOLYHEDROVIRUS± PARASITOID COMPETITION 655 TABLE 1. Viral occlusion body (OB) production in diverent instars of Spodoptera frugiperda larvae nonparasitized or parasitized by Chelonus insularis Production of virus at each instar a (Mean number OBs/ larva SE) Larval condition L 2 L 3 L 4 L 5 Nonparasitized a b c d Parasitized, C. insularis a b b ND ND: not determined as parasitoid emergence occurred in fth instar. a Means followed by the same letter are not signi cantly diverent, Fisher s LSD test (P > 0.05). Data were obtained as one piece of the information for each larva and, therefore, each data point was independent of the others. Pst I BglII kb M Wt A B C Wt D * * * * FIGURE 2. Restriction endonuclease pro les of viral DNA, ampli ed in Spodoptera frugiperda larvae nonparasitized, parasitized by Chelonus insularis following digestion with PstI or BglII. M (Marker), wt (wild-type virus), A, B and C are virus isolates that appeared in nonparasitized larvae, D is a virus isolate that appeared only in parasitized larvae. A HindIII digest of k DNA was used as molecular marker. Asterisks indicate the presence of distinct submolar fragments in each isolate.

8 656 A. ESCRIBANO ET AL. TABLE 2. Restriction endonuclease pro les generated after serial passage of wild-type (WT) virus in nonparasitized and parasitized Spodoptera frugiperda larvae Treatment and PstI pro le a BglII pro le a passage number I II III I II III Nonparasitized P 1 WT WT WT WT WT WT P 3 WT WT WT WT WT WT P 6 A C WT WT WT WT P 1 0 B C A WT WT WT Parasitized C. insularis P 1 WT WT WT WT WT WT P 3 WT WT WT WT D WT P 6 WT WT WT WT D WT P 1 0 WT WT WT WT D WT a WT, A, B, C, and D represent the diverent virus isolates identi ed by their DNA restriction endonuclease pro les presented in Figure 2. I, II and III represent the results of three replicate experiments. Southern blot analysis of the pro le D indicated that the submolar fragment at 14.5 kb derived from the viral genome, since hybridizatio n signals were detected in the DNA isolated from virus isolate D and from wild-type virus but not with the DNA isolated from nonparasitized S. frugiperda larvae, S. frugiperda larvae parasitized by C. insularis with or without the parasitoid larva, and adult female C. insularis (Figure 3). The probe showed FIGURE 3. (a) Restriction endonuclease pro les of BglII-digested DNA from: virus isolate D (lane 1), wild-type virus (lane 2), Spodoptera frugiperda larvae (lane 3), S. frugiperda parasitized larvae (lane 4) S. frugiperda parasitized larvae with parasitoid larva removed (lane 6) and adult female C. insularis (lane 5). (b) Southern blot analysis of gel shown in (A) using the 14.5 kb submolar BglII fragment of virus isolate D as probe. Lane k, lambda DNA digested with HindIII.

9 NUCLEOPOLYHEDROVIRUS± PARASITOID COMPETITION 657 strong hybridizatio n with the corresponding submolar fragment at 14.5 kb in the virus isolate D pro le but no hybridization was detected in the wild-type virus pro le at the 14.5 kb migration point indicating that the submolar fragment was not present in detectable quantities in the wild-type isolate. Hybridization signals of diverent intensities were also detected in various equimolar restriction fragments of both virus isolate D and wild-type virus DNA. Biological Activity of Virus in Serial Passage Experiment The infectivity of each virus isolate (WT, A, B, C, and D) was determined by bioassay and, in all cases, virus induced mortality increased with virus concentration. The probit analysis 2 regression line slopes were tted in parallel with a common slope of 0.76 (v , df 5 4, P ) (Table 3). The relative potencies of the four novel virus isolates was calculated as the ratio of the probit regression of the wild-type and each novel isolate, with the corresponding 95% ducial limits (Finney, 1971). Virus produced in parasitized larvae (virus isolate D) was 5.5 times more potent than the original WT inoculum. No signi cant diverences were found between the virus pro les A and B compared to wild type virus, while the virus isolate C was signi cantly less infective. The pro le isolates also divered in their virulence as determined by mean time to death of insects infected in bioassays; only isolate A was not signi cantly diverent from the wildtype virus. Of the remaining isolates, isolate D showed a 10.8% reduction in mean lethal time and isolate C was still more virulent, with a 16.8% reduction in mean lethal time compared to the wild type virus. Isolate B was intermediate between C and D (Table 3). DISCUSSION Both parasitism and viral infection resulted in a marked reduction in host growth. When third instar larvae were dually parasitized and infected, the growth rate was even more severely avected compared to parasitized larvae. There was a signi cant increase in virus production in larvae infected at later instars, re ecting the correlation between log body weight and log virus production reported in baculovirus infections of other species of Lepidoptera (Evans et al., 1981; Kunimi et al., 1996). Interspeci c competition for host resources resulted in a substantial decrease in pathogen production in parasitized larvae infected at the fourth instar, but not in parasitized larvae infected at earlier instars. This evect may be related to the stage of development of the parasitoid and its evects on the rate of growth of the host larvae; the yield of viral occlusion bodies is positively correlated with host growth rate (Shapiro, 1986). Almost no host growth was observed in dually parasitized and infected fourth instar larvae during the 96 h study period (Figure 1(b)) whereas the weight of parasitized + infected third instar larvae more than doubled in the same period (Figure 1(a)). The survival of C. insularis was not possible in S. frugiperda larvae that ingested a lethal dose of the nucleopolyhedroviru s during the second, third or early fourth larval instars con rming the results previously reported by Escribano et al. (2000). The serial passage experiment resulted in the appearanc e of distinct genetic isolates. Of the three isolates from nonparasitize d larvae, two showed increased virulence and for one of these (isolate C), the infectivity was reduced. One isolate arose in parasitized larvae which had increased virulence and infectivity. There are several possible explanations for the origins of these genetic isolates. DNA hybridizatio n analysis revealed that the 14.5 kb submolar fragment comprised virus genomic sequences; hybridization signals being detected in the BglII digest of virus isolate D and the wild-type isolate, but not in DNA from unparasitize d or parasitized S. frugiperda larvae (with or without the parasitoid larvae) or adult female C. insularis wasps. Since the Southern hybridizatio n experiment was performed three times, using a diverent probe preparation on each occasion, the most likely explanatio n for the hybridization signals detected in the wildtype isolate is that regions homologous to probe sequences exist in other parts of the virus

10 658 A. ESCRIBANO ET AL. TABLE 3. Response of second instar Spodoptera frugiperda larvae to wild-type virus and four virus isolates (A± D) harvested after serial passage in S. frugiperda larvae nonparasitized and parasitized by Chelonus insularis 95% CL 95% CL Virus Relative Mean time variant Regression equation SE of slope LC 5 0 (OBs ml - 1 ) potency lower upper to death (h) lower upper WT 5 3 Ð Ð y 0.76x A y 0.76x B y 0.76x C y 0.76x D y 0.76x Regression lines, LC5 0 values, and relative potency with respect to the wild-type (wt) virus, were calculated using the POLO-PC program (Le Ora Software, 1987). LC 5 0 values of virus isolates A and B did not diver signi cantly from the wt isolate, whereas isolates C and D were signi cantly less and more infective, respectively. The time-mortality data were subjected to analysis of variance using the generalized linear modelling package of the program SPSS (v. 7.5). Isolates A and B did not diver signi cantly from the wt isolate, whereas isolates C and D were signi cantly less and more virulent, respectively.

11 NUCLEOPOLYHEDROVIRUS± PARASITOID COMPETITION 659 genome. These results indicate that virus isolate D did not originate via insertion of a host or parasitoid transposon. It also appears unlikely that a latent virus infection was involved, as has been identi ed in a laborator y culture of Mamestra brassicae (Hughes et al., 1997), due to the fact that isolate D appeared only in parasitized hosts and not in unparasitized conspeci cs from the same culture. Parasitism, however, may have caused immunosupression as Chelonus species have an associated polydnavirus (Stoltz & Whit eld, 1992), and this might have permitted the expression of a latent virus, although we saw no signs of baculoviru s disease in parasitized hosts that were not inoculated with virus. An alternative explanation is that isolate D was generated by selection of a genotypic variant present in the wild-type isolate at a very low concentration. Natural baculoviru s isolates typically represent a mixture of individual genotypes that may be cloned in vitro by plaque puri cation (Maruniak et al., 1984; Ribeiro et al., 1997) or in vivo by limiting dilution assays in host larvae (Smith & Crook, 1988). These variant s can exist at such low concentrations that their presence cannot be detected using DNA hybridisation techniques (MunÄ oz et al., 1999). We suggest this possibility is also unlikely as isolate D appeared in only one of three replicate experiments using parasitized larvae run in parallel under the same conditions, and would therefore have experienced identical pressures of selection resulting in the appearance of this isolate in all three replicates. Finally, we suggest that isolate D was most likely generated by point mutation of a restriction site or by alterations such as duplications, deletions or by recombination of two or more genotypic variants present in the wild-type nucleopolyhedroviru s isolate. Each of these mechanisms has been identi ed in the formation of genotypic variants of the nucleopolyhedroviruse s of Anticarsia gemmatalis (Croizier & Ribeiro, 1992; Garcõ Â a-maruniak et al., 1996; Maruniak et al., 1999) and Spodoptera exigua (MunÄ oz et al., 1998; 1999). Genotypic variants present in one wild type SfMNPV isolate have been cloned and mapped (Maruniak et al., 1984) and many others have been detected by the presence of submolar bands in restriction endonuclease pro les (Shapiro et al., 1991; Escribano et al., 1999). This possible origin of virus isolate D is compatible with the observed results in that it may have only been generated in one of the three replicates, and due to its increased virulence, would have been selected favourably over other virus variants. DiVerences in infectivity or virulence between genetic variants present within an isolate have been reported before. Genotypic variant s cloned from wild-type A. gemmatalis nucleopolyhedroviru s (Ribeiro et al., 1997) exhibited marked diverences in infectivity while variants from Autographa californica nucleopolyhedroviru s appeared to be phenotypically uniform in terms of infectivity and virulence (Andrews et al., 1980). DiVerences in the infectivity of certain nucleopolyhedroviru s variants towards heterologous host species have also been reported (Weitzman et al., 1992; Garcõ Â a-maruniak et al., 1996). Recently, MunÄ oz et al. (1998) presented evidence that two genotypic variants present in a nucleopolyhedroviru s wild-type isolate were naturally-occurrin g deletion mutants. These variants were parasitic; for replication, they required the presence of genes present in other virus variants. Furthermore, the presence of deletion mutants was associated with a reduction in the insecticidal activity of the wild-type isolate and of other variant mixtures in which they were present. If genetic variants present within an isolate diver in their biological activity, they are likely to compete intraspeci cally within the host. Serial passage is likely to lead to dominance of whatever strain(s) are favoured by the prevailing conditions. For example, under eld conditions, serial passage of A. gemmatalis NPV (AgMNPV) over the course of six years by soya growers in Brazil resulted in an increase in genetic heterogeneity that was associated with greater heterogeneity in the virulence of variants cloned out of recycled inoculum (Maruniak et al., 1999). Abot et al. (1995) reported an overall decrease in viral infectivity following recycling of AgMNPV inoculum. In contrast, a high degree of stability of phenotypic traits has previously been reported following 20 serial passages of isolates of Helicoverpa zea NPV (McIntosh & IgnoVo, 1986). In the present study, no changes were detected in the infectivity or virulence of the wild-type SfMNPV material following ten

12 660 A. ESCRIBANO ET AL. passages, but in several replicates previously undetectable variants with divering phenotypes became more abundant. Our study provide s clear evidence of interspeci c competition within the host, since, depending on the timing of inoculation, adverse evects were found upon both the parasitoid and the virus. Direct consequences on parasitoids developing in virus-infected hosts have been reported including the production of virus-induced toxic factors that inhibit parasitoid development (Kaya & Tanada, 1971, 1973; Hotchkin & Kaya, 1983), premature death of the host (Brooks, 1993) and the preclusion of parasitoid persistence in virus-infected host populations (Bird & Elgree, 1957). The body size of a braconid endoparasitoid was reduced in Heliothis virescens larvae infected with a recombinant baculovirus expressing an insect neurotoxin gene, probably due to a reduced development time in recombinant virus infected hosts compared to larvae infected with the wild-type virus (McCutchen et al., 1996). Conversely, the presence of a developing parasitoid may have detrimental evects on the production of progeny virus (Caballero et al., 1990; and this study), but parasitoids that emerge from baculovirus-infecte d hosts have also been shown to be responsible for virus transmission during the act of ovipositio n (Young & Yearian, 1990; Hochberg, 1991b; Caballero et al., 1991) and as such, parasitoids may be important agents for the dispersal of these viruses. This study presents the rst evidence of the consequences of parasitism on characteristics pivotal to the insecticidal potential of the virus. The virus isolate that appeared during serial passage of virus exclusively in parasitized hosts showed increase virulence and infectivity compared to wild-type virus. It seems unlikely that this was a purely chance observation as none of the three isolates that appeared in non-parasitize d hosts were observed in dually parasitized and infected larvae. Within-host intraspeci c competition in serial passage experiments drives the evolution of virulence; the pathogen strain with the highest replication rate will achieve the greatest numerical representation in the inoculum used to infect the following generation of insects and will thereby have a selective advantag e over competitor strains (Ebert, 1998). In parasitized hosts infected by virus, the situation is slightly more complicated as all strains of the pathoge n also experience interspeci c competition for host resources with the developing parasitoid larvae. As yet, we understand very little about the processes that maintain genetic and phenotypic diversity within pathogen populations. Presumably, changes in virus infectivity and virulence carry associated costs in some other component(s) of virus tness but for the moment these remain unknown. ACKNOWLEDGEMENTS This research was funded by the European Commission (contract number IC18-CT96± 0097). A. Escribano received a grant from the Universidad Pu blica de Navarra. T. William s received additional support from SIBEJ REFERENCES Abot, A.R., Moscardi, F., Fuxa, J.R. Sosa-Go mez, D.R. & Richter, A.R. (1995) Susceptibility of populations of Anticarsia gemmatalis from Brazil and the United States to a nuclear polyhedrosis virus. Journal of Entomological Science 30, 62± 69. Andrews, R.E., Spence, K.D. & Miller, L.K. (1980) Virulence of cloned variants of Autographa californica nuclear polyhedrosis virus. Applied and Environmental Microbiology 39, 932± 933. Ashley, T.R. (1986) Geographical distributions and parasitization levels for parasitoids of the fall armyworm, Spodoptera frugiperda. Florida Entomologist 69, 516± 524. Beegle, C.C. & Oatman, E.R. (1974) DiVerential susceptibility of parasitized and non-parasitized larvae of Trichoplusia ni to a nuclear polyhedrosis virus. Journal of Invertebrate Pathology 24, 188± 198. Bird, F.T. & Elgree, D.E. (1957) A virus disease and introduced parasites as factors controlling the European spruce saw y, Diprion hercyniae (Htg.) in Central New Brunswick. Canadian Entomologist 89, 371± 378.

13 NUCLEOPOLYHEDROVIRUS± PARASITOID COMPETITION 661 Blissard, G.W., Black, B.N., Crook, N.E. Keddie, R., Possee, R.D., Rohrmann, G.F., Thielmann, D.A. & Volkman, L.E. (2000) Baculoviridae, in Virus Taxonomy: Seventh Report of the International Committee on Taxonomy of Viruses (Van Regenmortel, M.H.V., Fauquet, C.M., Bishop, D.H.L., Carstens, E.B., Estes, M.K., Lemon, S.M., Maniloff, J., Mayo, M.A., McGeoch, D.J., Pringle, C.R. & Wickner, R.B., Eds). Academic Press, New York, pp. 195± 202. Bonning, B.C. & Hammock, B.D. (1996) Development of recombinant baculoviruses for insect control. Annual Review of Entomology 41, 191± 210. Boots, M. & Begon, M. (1995) Strain diverences in the Indian meal moth, Plodia interpunctella, in response to a granulosis virus. Researches on Population Ecology 37, 37± 42. Brooks, W.M. (1993) Host-parasitoid-pathogen interactions, in Parasites and pathogens of insects, Vol. 2: Pathogens (Beckage, N.E., Thompson, S.N. & Federici, B.A., Eds). Academic Press, San Diego, pp. 231± 272. Caballero, P., Vargas-Osuna, E. & Santiago-Alvarez, C. (1990) Development of Apanteles telengai (Hym.: Braconidae) and Campoletis annulata (Hym.: Ichneumonidae) in granulosis virus (GV) infected Agrotis segetum (Lep.: Noctuidae) larvae. Journal of Applied Entomology 110, 358± 364. Caballero, P., Vargas-Osuna, E. & Santiago-Alvarez, C. (1991) Parasitization of granulosis-virus infected and noninfected Agrotis segetum larvae and the virus transmission by three hymenopteran parasitoids. Entomologia Experimentalis et Applicata 58, 55± 60. Caballero, P., Zuidema, D., Santiago-Alvarez, C. & Vlak, J.M. (1992) Biochemical and biological characterization of four isolates of Spodoptera exigua nuclear polyhedrosis virus. Biocontrol Science and Technology 2, 145± 157. Croizier, G. & Ribeiro, C.P. (1992) Recombination as a possible major cause of genetic heterogeneity in Anticarsia gemmatalis nuclear polyhedrosis virus populations. Virus Research 26, 183± 196. Ebert, D. (1998) Experimental evolution of parasites. Science 282, 1432± Eller, F.J., Boucias, D.G. & Tumlinson, J.H. (1988) Interactions between Microplitis croceipes (Hymenoptera: Braconidae) and a nuclear polyhedrosis virus of Heliothis zea (Lepidoptera: Noctuidae). Environmental Entomology 17, 977± 982. Escribano, A., Williams, T., Goulson, D. Cave, R.D., Chapman, J.W. & Caballero, P. (1999) Selection of a nucleopolyhedrovirus for control of Spodoptera frugiperda (Lepidoptera: Noctuidae): structural, genetic, and biological comparison of four isolates from the Americas. Journal of Economic Entomology 92, 1079± Escribano, A., Williams, T., Goulson, D. Cave, R.D. & Caballero, P. (2000) Parasitoid-pathogen-pest interactions of Chelonus insularis, Campoletis sonorensis, and a nucleopolyhedrovirus in Spodoptera frugiperda larvae. Biological Control 19, 265± 273. Evans, H.F., Lomer, C.J. & Kelly, D.C. (1981) Growth of nuclear polyhedrosis virus in larvae of the cabbage moth, Mamestra brassicae L. Archives of Virology 70, 207± 214. Finney, D.J. (1971) Probit Analysis. Cambridge University Press, London, UK. Garci a-maruniak, A., Pavan, O.H.O. & Maruniak, J.E. (1996) A variable region of Anticarsia gemmatalis nuclear polyhedrosis virus contains tandemly repeated DNA sequences. Virus Research 41, 123± 132. Heckel, D.G., Gahan, L.J., Tabashnik, B.E. & Johnson, M.W. (1995) Randomly ampli ed polymorphic DNA diverences between strains of diamondback moth (Lepidoptera; Plutellidae) susceptible or resistant to Bacillus thuringiensis. Annals of the Entomological Society of America 88, 531± 537. Hochberg, M.E. (1991a) Intra-host interactions between a braconid endoparasitoid, Apanteles glomeratus, and a baculovirus for larvae of Pieris brassicae. Journal of Animal Ecology 60, 51± 63. Hochberg, M.E. (1991b) Extra-host interactions between a braconid endoparasitoid, Apanteles glomeratus, and a baculovirus for larvae of Pieris brassicae. Journal of Animal Ecology 60, 65± 77. Hotchkin, P.G. & Kaya, H.K. (1983) Pathological response of the parasitoid, Glyptapanteles militaris, to nuclear polyhedrosis virus-infected armyworm hosts. Journal of Invertebrate Pathology 42, 51± 61. Hughes, P.R. & Wood, H.A. (1981) A synchronous peroral technique for the bioassay of insect viruses. Journal of Invertebrate Pathology 37, 154± 159. Hughes, D.S., Possee, R.D. & King, L.A. (1997) Evidence for the presence of a low-level, persistent baculovirus infection of Mamestra brassicae insects. Journal of General Virology 78, 1801± Kaya, H.K. & Tanada, Y. (1971) Properties of a viral factor toxic to the parasitoid Apanteles militaris. Journal of Insect Physiology 17, 2125± Kaya, H.K. & Tanada, Y. (1973) Hemolymph factor in nuclear polyhedrosis virus toxic to Apanteles militaris. Journal of Invertebrate Pathology 21, 211± 214. Kolodny-Hirsch, D. & Van Beek, N.A.M. (1997) Selection of a morphological variant of Autographa californica nuclear polyhedrosis virus with increased virulence following serial passage in Plutella xylostella. Journal of Invertebrate Pathology 69, 205± 211. Kunimi, Y., Fuxa, J.R. & Hammock, B.D. (1996) Comparison of wild type and genetically engineered nuclear polyhedrosis viruses of Autographa californica for mortality, virus replication and polyhedra production in Trichoplusia ni larvae. Entomologia Experimentalis et Applicata 81, 251± 257. Le Ora Software (1987) POLO-PC a User s Guide to Probit or Logit Analysis. Berkeley, CA, USA. Marti nez, A.M., Goulson, D., Chapman, J.W. Caballero, P., Cave, R. D. & Williams, T. (2000) Is it

14 662 A. ESCRIBANO ET AL. feasible to use optical brightener technology with a baculovirus bioinsecticide for resource-poor maize farmers in Mesoamerica? Biological Control 17, 174± 181. Maruniak, J.E., Brown, S.E. & Knudson, D.L. (1984) Physical maps of SfMNPV baculovirus DNA and its genomic variants. Virology 136, 221± 234. Maruniak, J.E., Garcia-Maruniak, A., Souza, P. Zanotto, M.A. & Moscardi, F. (1999) Physical maps and virulence of Anticarsia gemmatalis nucleopolyhedrovirus genomic variants. Archives of Virology 144, 1991± McCutchenn, B.F., Herrmann, R., Heinz, K.M. Parrella, M.P. & Hammock, B.D. (1996) EVects of recombinant baculoviruses on a nontarget endoparasitoid of Heliothis virescens. Biological Control 6, 45± 50. McIntosh, A.H. & Ignoffo, C.M. (1986) Restriction endonuclease cleavage patterns of commercial and serially passaged isolates of Heliothis Baculovirus. Intervirology 25, 172± 176. Milks, M.L. (1997) Comparative biology and susceptibility of cabbage looper (Lepidoptera: Noctuidae) lines to a nuclear polyhedrosis virus. Environmental Entomology 26, 839± 848. Miller, L.K. & Lu, A. (1997) The molecular basis of baculovirus host range, in The Baculoviruses (L.K. Miller, Ed.). Plenum Press, New York, pp. 217± 236. Moscardi, F. (1989) Use of viruses for pest control in Brazil: the case of the nuclear polyhedrosis virus of the soy bean caterpillar, Anticarsia gemmatalis. Memorias Instituto Oswaldo Cruz 84, 51± 56. Moscardi, F. (1999) Assessment of the application of baculoviruses for control of Lepidoptera. Annual Review of Entomology 44, 257± 289. Moscardi, F., Leite, L.G. & Zamataro, C.E. (1997) Production of nuclear polyhedrosis virus of Anticarsia gemmatalis HuÈ bner (Lepidoptera: Noctuidae): evect of virus dosage, host density and age. Annals of the Entomological Society of Brasil 26, 121± 32. MunÄ oz, D., Vlak, J.M. & Caballero, P. (1997) In vivo recombination between two strains of the genus Nucleopolyhedrovirus in its natural host Spodoptera exigua. Applied and Environmental Microbiology 63, 3025± MunÄ oz, D., Castillejo, J.I. & Caballero, P. (1998) Naturally occurring deletion mutants are parasitic genotypes in a wild-type nucleopolyhedrovirus population of Spodoptera exigua. Applied and Environmental Microbiology 64, 4372± MunÄ oz, D., Murillo, R., Krell, P.J. Vlak, J.M. & Caballero, P. (1999) Four genotypic variants of a Spodoptera exigua nucleopolyhedrovirus (Se-SP2) are distinguishable by a hypervariable genomic region. Virus Research 59, 61± 74. Pavan, O.H.O. & Ribeiro, H.C.T. (1989) Selection of a baculovirus strain with a bivalent insecticidal activity. Memorias Instituto Oswaldo Cruz 84, 63± 65. Ribeiro, H.C.T., Pavan, O.H.O. & Muotri, A.R. (1997) Comparative susceptibility of two diverent hosts to genotypic variants of the Anticarsia gemmatalis nuclear polyhedrosis virus. Entomologia Experimentalis et Applicata 83, 233± 237. Sambrook, J., Fritsch, E.F. & Maniatis, T. (1989) Molecular Cloning: a Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, USA. Santiago-Alvarez, C. & Caballero, P. (1990) Susceptibility of parasitized Agrotis segetum larvae to a granulosis virus. Journal of Invertebrate Pathology 56, 128± 131. Shapiro, D.I., Fuxa, J.R., Braymer, H.D. & Pashley, D.P. (1991) DNA restriction polymorphism in wild isolates of Spodoptera frugiperda nuclear polyhedrosis virus. Journal of Invertebrate Pathology 58, 96± 105. Shapiro, M. (1986) In vivo production of baculoviruses, in The Biology of Baculoviruses. Vol 2: Practical Application for Insect Control (Granados, R.R. &. Federici, B.A., Eds). CRC Press, Florida, pp. 31± 62. Smith, I.R.L. & Crook, N.E. (1988) In vivo isolation of baculovirus genotypes. Virology 166, 240± 244. Stark, D.M., Mills, N.J. & Purcell, A.H. (1999) Interactions between the parasitoid Ametadoria misella (Diptera: Tachinidae) and the granulovirus of Harrisina brillians (Lepidoptera: Zygaenidae). Biological Control 14, 146± 151. Stoltz, D.B. & Whitfield, J.B. (1992) Viruses and virus-like entities in the parasitic Hymenoptera. Journal of Hymenoptera Research 1, 125± 139. Vasconcelos, D.S., Williams, T., Hails, R.S. & Cory, J.S. (1996) Prey selection and baculovirus dissemination by carabid predators of Lepidoptera. Ecological Entomology 21, 98± 104. Weitzman, M.D., Possee, R.D. & King, L.A. (1992) Characterization of two variants of Panolis Xammea multiple nucleocapsid nuclear polyhedrosis virus. Journal of General Virology 73, 1881± Wheeler, G.S., Ashley, T.R. & Andrews, K.L. (1989) Larval parasitoids and pathogens of the fall armyworm in Honduran maize. Entomophaga 34, 331± 340. Williams, T., Goulson, D., Caballero, P. Cisneros, J., Mati nez, A.M., Chapman, J.W., Roman, D.X. & Cave, R.D. (1999) Evaluation of baculoviruses bioinsecticide for small-scale maize growers in Latin America. Biological Control 14, 67± 75. Young, S.Y. & Yearian, W.C. (1990) Transmission of nuclear polyhedrosis virus by the parasitoid Microplitis croceipes (Hymenoptera: Braconidae) to Heliothis virescens (Lepidoptera: Noctuidae) on soybean. Environmental Entomology 19, 251± 256.